Which Sony Alpha?

Now that there are five Sony Alpha DSLR bodies in circulation, with many owners of the original 2006 Alpha 100 considering a replacement, the differences between this ur-Alpha and the 2007-8 generation of Alpha 700, 200, 300 and 350 need examining.

Price is one easy determining factor when choosing a DSLR, but even this has become muddied by exclusive deals, selected bodies available only with the 18-70mm kit lens, cashback offers and the whole gamut of retail promotions and packages. The Alpha 700 remains in a price-bracket of its own but the A200, 300 and 350 are clustered within a tight range. In the UK, you can spend between £300 and £400 and get whichever body you want, depending on whether you need a kit lens or not.

The Alpha 700

The Alpha 700 is simple to look at first. It is currently available in the UK for under £750 body only, including a £100 cash-back deal from Sony which has been extended indefinitely as far as can be told, and includes 17.5% VAT. This pitches it slightly higher than the Canon 40D (which is only 10 megapixels compared to the A700’s 12) and below the Nikon D300.

The 700 has a slightly larger body than the 200-350 series, clearly based on the earlier Konica Minolta 7D. It is not just a matter of size; the camera is built using magnesium alloy shell components over a tempered aluminium framework, skinned in a durable and resilient polymer. It has a true glass pentaprism with a generous viewfinder magnification, giving both bright and comfortable focusing screen viewing. The rear LCD is the largest (3 inches) and highest resolution (930,000 pixels) of any fitted to any DSLR made (it is shared with the Nikon D300, as is its 12.2 megapixel Sony CMOS sensor).

These factors combine with what is probably the best user interface ever designed for a DSLR, Sony’s Quick Navi. You might imagine that this extremely fast and clear GUI would have been used on the later 200-350, but that is not so. The more ‘professional’ Alpha 700 actually needs fewer button presses or user selection actions to make changes when shooting, and is more intuitive.

It’s hard to describe the difference in the shooting experience between the earlier Alpha 100, new 200-350 series and the 700. It has nothing to do with automation or shortcuts as it is faster when working entirely manually as well, due to the use of twin control wheels front and rear (finger and thumb). It’s a bit like opting for a generously sized and powered high end German car in place of a modest sized budget French runabout – as much to do with comfort, feel and tactile feedback as with the ability to do 180kph down the autobahn.

However, speed is one of the points of the Alpha 700. It will shoot sequences at a nominal 5 frames per second (tested by us at a practical 4.5-4.8fps), taking the fastest speed to 1/8000th and flash sync up to 1/250th; it has ISO settings up to 6400, a focusing system with a special sensor to make full use of lense faster than f/2.8, and the fastest most accurate autofocus sensor and AF-motor assembly. It writes to CF cards making full use of UDMA specifications but also downloads via its USB 2.0 connection at higher speeds, and it is one of the few DSLRs made which will drive a 1080p HD-TV at full resolution.

To enable HD-TV playback it also comes with an infra-red remote hand controller, which further allows normal and 2-second self timer wireless shutter release with autofocus. This accessory would have a street price close to £50 and that should be considered when comparing the price with the 200-350 models, as they lack this controller or the ability to communicate with it.

Having graduated from the Alpha 100 first to the Alpha 700, and later acquired Alpha 200 and 350 models, the only thing which prevents us here at photoworldalpha disposing of the Alpha 200 and both earlier 100s to acquire a second 700 body is a need to write articles like this. Without the earlier or lesser bodies to hand, comparisons are soon forgotten and all we have is an archive of images to study. Often I need the cameras to hand to answer calls from Photoworld Club members.

There is no doubt after a reasonable period of using the Alpha 700, followed by some time with the later cameras, that it’s the model to own if you can justify the outlay. One minor point which has even been criticised by some reviewers – the inclusion of a second memory slot accepting Sony/SanDisk Memory Stick Pro Duo cards – is important for any professional user. During testing the Alpha 350, and continuing to use the Alpha 700 as normal, I picked up what I thought was the 700, which I had just loaded with a new CF card. In fact I’d got the 350, which had no card in it at all!

Even if that happened with the 700, I have a 2GB card permanently in the second slot. It is never used, but the day I pick up the camera and forget to load a CF card, or run out of card space unexpectedly, I have a built-in standby to save my skin. It’s in there. You just wouldn’t know it!

Alpha 700 image quality

It is eight months since I acquired the Alpha 700, and many of our readers have asked – when will you publish your report on Alpha 700 image quality? The short answer is that image quality is relative, and in the subsequent period, I have been bombarded with DSLRs to review for the British Journal of Photography. The Olympus E-3, Nikon D300, Nikon D3, Canon 40D (before the 700), Pentax K20D, Nikon D60 and Canon 450D have all been through my hands.

There is only one of these cameras I would rather use than the Alpha 700, and that’s the Nikon D3. There is one I would be equally happy with, the Pentax K20D – and I would extend that to the Nikon D300, which is effectively an Alpha 700 in a Nikon body with a few enhancements, but for one missing feature. In-body image stabilisation!

One tenth of a second hand-held with the Alpha 700 and 11-18mm lens at 11mm, f/8.

A 100 per cent clip from this shot – SSS is a great asset in the Sony DSLRs.

My dilemma in writing up the image quality of the Alpha 700 is that theoretical image quality and real life quality are two different things. My normal IQ test, using studio flash and a top grade 50mm lens (macro, standard, or a premium zoom in the absence of a prime) always shows very small differences between cameras of similar pixel count when used at minimum or optimum ISO. Most of this can be put down to focus calibration differences, which affect all makes and all levels of body from entry to top pro.

However, in the real world exposures are not made on a tripod with flash at 1/1500th of a second. Nor are most of them made at borderline hand-holding speeds like 1/30th with a 50mm lens or 1/125th with a 300mm. Plenty are made with shutter speeds you do not think are marginal – say 1/250th with a lens set to 300mm. Safe? Not so. A 300mm lens on the 1.5X format matches using a 450mm on full frame 35mm, and that would call for 1/500th as the SLOWEST hand-holdable speed working on the old formula of one-over-the-focal length.

That formula was devised by the early users of Leica rangefinder cameras, Rollei 6 x 6 TRLs and similar kit. They were generally aiming to make a sharp 10 x 8 inch print, and even the depth of field tables were worked out on that basis. My 24 inch iMac screen is almost four times the size of the 6 x 9 inch image area I used to print on 10 x 8 paper. My Epson 3800 makes great 13 x 19 inch prints, or bigger. Once, a 20 x 16 print was something you made only from the best dozen images shot in a year – a special enlargement, for exhibition or entry into competitions. Today we view and print pictures this size casually.

I can not emphasise how useful it is to keep SSS (Super Steady Shot) active even when you know the shutter speed will be well into three figures (over 1/100th, which many shooters feel is ‘safe’ regardless of the lens in use). On a few occasions I have accidentally disabled SSS, failing to turn it back on after a studio shoot, and lost the ‘edge’ I am used to seeing from my Alpha 700 as a result.

You could mistake the difference for a poorer sensor, poor focusing or a cheap lens but the real cause will be the absence of the in-body stabilisation which gives every single shot taken an advantage, no matter what lens you are using. Now this applies across the entire Alpha range, and having used two alternative in-body systems, I believe the current Sony implementation is by far the best at those ‘false sense of security’ shutter speeds from 1/30th all the way up to 1/1000th.

Image quality with the Alpha 700 is down to two factors – your choice of lens, and the ISO/NR combination or related image processing. I always shoot raw even if also saving JPEGs for reference, and I process the files using a range of converters but mainly Adobe Camera Raw and Lightroom. These both share a core engine and a reputation for getting Alpha 700 files seriously wrong. There is no doubt that they do, and the error is not in the NR or de-Bayer processing, it’s in the contrast curve and black point applied by default. I correct that situation by creating new camera raw defaults, changing the Brightness level to 0 and the Black Point to 0 as a startpoint.

Processing Canon 450D and Alpha 700 raw files side by side, the 450D has a slight noise advantage from ISO 400 to 1600. Since the Canon is limited to 1600, comparisons at 3200 and 6400 are impossible. Where the Alpha 700 loses is in a certain type of midtone which is often found in defocused areas. It actually has better shadow noise than the 450D, and underexposed images shot at ISO 400-800 can be pushed in raw development. Present both cameras with a well-exposed lighter toned area full of detail, and both will show hardly any noise even at 800 in the light detail.

It seems that Sony has applied noise reduction most to the darker tones, leaving the fairly steep midtone curve to its own devices. The downside of this is that one particular set of tones – sky blues – fall right into the noisiest zone of the Alpha 700’s response even at IS0 100-400. The red channel seems to suffer most from grainy noise and the blue channel least. Skies in Alpha 700 pictures can be more noisy than you would want, and more noisy than other makes (this noise is caused by the red and green channels, not by the blue itself).

The worst comparison is with the Pentax K20D, which suppresses midtone noise most effectively. It does not have the same reserve of dynamic range into the shadows and highlights that the A700 does (this is also true of the Canon 450D) and here, I think, lies the answer to why Sony images seem to have more ‘life’ and also higher in midtone noise. If you reserve more of the available bit depth to handle highlight recovery (each stop at the highlight end uses half of any remaining data depth), and also try to recover the maximum shadow detail, you end up with fewer tone-steps available in the critical mid range. If you then apply a natural-looking tone slope to the mid range, you increase the contrast between the visible steps.

This is what I’m looking for in my Sony A700 images – the punch to survive in on-line photo libraries, display well on screen, and make great prints. But blue skies like this can show some noise.

The result is part of the unique Sony ‘look’ – images which have a natural visual brightness and colour, but also have good scope for shadow adjustment and highlight recovery. This type of contrast handling is vital for effective implementation of the DRO+ (Dynamic Range Optimizer) because without a reserve of image information at both ends of the tonal scale, DRO+ has nothing to work on. And if the image does not already have a good bright visual contrast, DRO+ will make it seem artificially compressed and flat.

Let’s just say that while both the Pentax K20D and the Canon 450D bettered the noise-levels of the Alpha 700 across their ranges, both had disappointly muted and soft looking images to go with this. Flattening off the Sony image to resemble the Canon default immediately lowered the apparent noise; tweaking a Canon image to have the same vibrant colour and contrast as the Sony default increased its noise. The same applies to the Pentax, not quite as markedly, and considering the 14.6 megapixel count of that camera, they have put Canon’s claim to the low noise crown to the test and won.

It seems that noise and ISO and sharpness and saturation, together with what reserves of usable shadow and highlight recovery the raw file contains, may depend on a complex set of relationships which apply more equally to all makers than we think. Part of Canon’s anti-noise strategy depends on a gentler contrast slope and stronger anti-aliasing, resulting in an image which looks soft but smooth. Sony’s images (in general) appear more detailed and punchy, resulting in more noise. If Sony makes any error it is to apply in-camera NR once this stage has been reached, resulting in JPEGs where the detail is smoothed along with the noise. The effect is ragged compared to the best alternative methods. You can avoid strong noise reduction by shooting at ISO 1250 instead of 1600, just below the threshhold at which mandatory NR starts to happen in-cameras.

The final test – to print images from these different cameras – nearly always favours the Alpha 700, and as we shall see, its newer brothers. That’s because the JPEGs, or the default raw conversions, simply have better contrast and colours for printing.

You will notice that I have left Nikon out of this. There’s a good reason. Nikon’s competing cameras actually use the Sony sensors, and in the D60 especially you can observe a similar strategy for colour and tone rendering to the Alpha 200-300. If anything Nikon go even further towards a ‘snappy’ result in the D60 (the D300 is more restrained). Yet Nikon has found a way to keep strong midtone contrast without the same levels of noise.

Their solution is rather disarming in its simplicity. Let’s say the Alpha 700 is set to ISO 400, and aimed at a scene. The camera exposes for 1/500th at f/10, and you get a normal looking JPEG. The Nikon aimed at the same scene with a similar lens, also set to ISO 200, exposes for 1/320th at f/10 – and the JPEG loooks pretty much the same. Does that mean the Nikon was really at ISO 250?

No. There is no such thing as an ISO rating for a raw file, and you can expose for a raw file with higher or lower average values, then convert accordingly. By studying the shadow detail and tone values, and typical exposures of Nikon DSLRs, I’ve concluded that Nikon generally allows more light to hit the sensor in any given situation, ISO for ISO, while also discarding some low-bit data that otherwise can contribute to noise. This all results in more generous exposure times or wider apertures and makes using stabilised lenses, and good quality glass, important.

My final conclusion has been that I simply get better, more useful final images from the Alpha 700 over a wide range of conditions – even though in some situations other cameras produce a superior result. The dynamic range of the Alpha 700 raw file makes single exposures good enough for pseudo-HDR processing and almost eliminates the need to bracket, ever. I get a quality of tone and colour which works well on screen and in print, and is very saleable. I regret it every time I leave the Alpha 700 behind because another camera needs testing, and return to find I do not have the same satisfaction in the results.

The Alpha 200 and 350

This brings me to the younger brothers of the Alpha 700, because in both cases I have felt obliged to leave the 700 at home and travel with just the newer, more consumer-orientated models. Have I felt the same dissappointment?

Well, the Alpha 200 is easy to deal with if you are an Alpha 100 owner. There is no contest, and the gain in quality per ISO step (at least over the 200 setting) is greater than the mere half-stop suggested by some reviewers. Against the 700, the comparison is difficult. In some ways the absence of a strong NR reduction strategy in the 200 makes its 1600 and 3200 files a better startpoint for post-processing, but the real strength of the camera lies in a clean ISO 1600 even in low light situations. The Alpha 100 has always had a surprisingly good ISO 1600 for daylight sports/action shooting, it just broke down with longer exposures and darker details.

The 200 is a significant improvement on the A100 design in some ways, but actually goes backwards in others. The interface requires more button presses and menu diving, many functions are missing (depth of field button, mirror lock up, etc) and some feel the changes to the pop-up flash design made to match the A300/350 reduce its functionality.

Our Alpha 100 and 200 bodies went to the US for a trip with our daughter and her partner. They used high ISOs more freely than I would have done, resulting in a few cases of uncancelled ISO 800 for outdoor sunny views, and plenty of ISO 400 in all conditions. Here, the difference between the 100 and 200 raw files was very noticeable. Hardly any ISO 400 to 800 Alpha 100 images were suitable for our image library, but all the 400s from the Alpha 200 were fine, and a few 800s processed up to look normal.

The ceiling inside the dome of the Capitol, Washington DC, by Ailsa Kilpatrick – Alpha 200 with 18-250mm Tamron zoom, 1/80th at f/6.3, 35mm focal length, hand-held, ISO 800.

A 100 per cent clip from a 5120 pixel wide picture library export from the raw file.

I turned a couple of the Alpha 100 high ISO shots into graphic grainy black and whites for our library stock, as they had no viable use in their original colour form. Nothing like that was necessary with the Alpha 200 work.

The Alpha 200 consistently underexposed in terms of raw file density, no bad thing as its highlight recovery potential is not as good as the Alpha 700 or (as far as I can tell) the 100. My guess is that some of the extra light-gathering capacity of the redesigned 10.2 megapixel CCD is used to provide a more generous actual exposure which is converted using lower A to D gain. The Alpha 100 images, even when working from raw files only, were a bit more contrasty and bright and I’d say Sony has learned something from Canon’s approach of flattening things down a bit.

This produces the impression, when examining A200 images, that they are a touch less sharp and detailed than the original A100. In fact the resolution seems to be similar, it’s just a matter of contrast affecting how we see the result. Sony seems to have used the available signal-to-noise ratio and bit depth of its 12-bit BIONZ processing to provide a more acceptable level of image noise at ISO 200-1600, as well as to add the high 3200 setting.

When working on some A200 raw files and noticing an apparent slight magenta cast in ACR 4.4, the JPEG was referred to and the same cast was present in that. No such cast seemed to be present in either A100 or A350 files. The results from the A200 are certainly very warm in colour and most people will find this improves portraits, even if it can make some landscapes a bit less verdant than they should be.

Like the Alpha 700, the Alpha 200 has fairly noisy sky blues once you go above ISO 200, and will occasionally produce noise at lower settings. The entire remainder of a scene can be noise-free yet the sky will be grainy. Compared to Canon and Pentax, this is a negative aspect of Sony image quality across the board and one which users notice easily. Some real effort needs to be put in to taming this colour-specific noise.

The Alpha 200 image quality is affected by a slight optimism from the AF system. Where the new faster AF of the 700 is also extremely precise, the AF of the 200 is just good enough. It’s fast – 1.7X faster than the Alpha 100 – but the sensor array does not seem to have been improved accordingly. It was too ready to lock focus when a tiny nudge of the setting would have been spot on. There was no serious back or front focus.

When the focus was perfectly nailed, the results could be exceptional and occasionally beat the higher resolution Alpha 350 on similar subjects. The 10.2 megapixel sensor is well matched to popular lenses and the 18-250mm Tamron which we use gains nothing by swapping it to the Alpha 350. The 16-80mm Carl Zeiss, in contrast, shows exactly the gain in detail you would expect from 14.2 million pixels versus 10.2. The 16-105mm Sony, a recent addition, behaves more like the Tamron; it’s happy on 10.2 megapixels, better on 12.2, then doesn’t show any further gain on 14.2.

The Alpha 350 replaced my 700 for a week shooting around 500 final stock images of Grand Canaria (a week spent driving from location to location on some interesting roads, including a few which were on the map, on the sat-nav but turned out to be foot and horse paths in real life). This was a mixed experience.

The Alpha 350 adds to the A200 design a Live View system based on a CCD camera in the prism viewing the focusing screen and relaying this to a 2.7 inch rear LCD on a tilting and extended articulated frame. To fit the hardware, the hollow mirror prism is reduced in size and the finder only has a 0.74X magnification, compared to 0.83X for the A200 and 0.87X for the A700. This leaves you viewing a reduced-sized window which is not to bad with long lenses. I’m a wide-angle shooter much of the time, and found the small view forced me to look at the actual scene more often to check details.

During the week, the Live View was only useful for half a dozen shots out of hundreds of situations. I was not setting out to find uses for it, I was working with the camera and only used it when needed. Those few pictures would have been harder or impossible without it, and included shots of people with the dog statues in Las Palmas cathedral square, where aware subjects would have spoiled the result.

The change to the design of the rear of the camera moves the four-way focus controller/selector pad over to the right. I was having problems with the A350 refusing to autofocus, and could not work out why. Eventually I found the reason. It is very easy, with the camera hand in the right hand for a single-handed shot (or otherwise), to let the base of the thumb depress the right hand selector of the pad. If this is held in continuously, the AF is disabled even when first pressure is taken on the shutter release. Lightly touching the pad, and keeping the pressure there, produces exactly the problem I was occasionally having. The cramped design of the right-hand end of the camera is just the wrong shape for my grip where the Alpha 700 is perfect.

Focus accuracy was no better than the A200, and any small errors are magnified. While our 11-18mm Sony proved perfectly up to provided sharpness for 14.2 megapixels, the A350 often failed to refocus it at all, and to get a sharp AF pic I would have to aim it at the ground near me, then back at my distant subject. Otherwise, if the lens had been focused on 15 feet, the A350 would think it really didn’t need refocusing at all for 12 feet or 30 feet. This was most noticeable when comparing the 18-250mm at 18mm and f/3.5 with the 11-18mm at 18mm and f/5.6 (their maximum apertures at this focal length). The A350 seemed to think there was enough depth of field with the 11-18mm, no need to focus any more accurately. The expected Tamron 10-24mm f/2.8-4 lens should solve this problem.

As for the image quality, my first impressions were that the contrast and saturation really had been dulled down to keep noise levels low on the 14.2 megapixel Alpha 350. Combined with a hint of underexposure, pictures could be like the Canon 400D and 450D and not much like a ‘default’ Sony DSLR. Despite this, there was the usual blue sky noise even at ISO 100. Once again, the entire image could be apparently smooth and noise-free but grey or blue skies showed noise.

After a few days I realised that the Alpha 350 has an incredible dynamic range. It defies the laws of physics to have more usable shadow detail and highlight reserve than the 700 or 200/300. The flat, almost linear contrast could be disappointing when reviewing images on the rear screen (they actually look better during live preview) and a typical exposure in sunshine was 1/320th at f/11 for ISO 100. That is actually the ‘correct’ exposure for ISO 200, not 100.

Then I found that the real noise levels of the A350 at IS0 400 beat all the other Alphas including the 700 – and that overexposed highlights can be pulled by two stops in ACR without colour clipping – and that overexposed pale blue skies show no noise – and that I had been accidentally reviewing my shots at 6144 pixel wide (25 megapixel) export size in ACR without realising it…

In short, the Alpha 350 began to pose a problem. The viewfinder is small; the flash raises up so low even a lens without a hood can cast a shadow; it’s slow at only 2fps and it eats card space; it’s an amateut body. But the results I was seeing in my normal working ISO range, 100-400, were showing up not only the 10.2 megapixel sensor but my 700 as well. I was using the Pentax K20D alongside the A350, and this also surprised me for similar reasons. A quantum leap seems to have been made in controlling noise with both sensors, yet they are entirely different – Sony CCD versus Pentax CMOS.

This left the big question – how about 800, 1600 and 3200?

Canarian guitarra, laud and timple on a bar wall in Fataga – ISO 800, details below.

This clip in not at 100 per cent – it’s from a 5120 pixel high, 17.5 megapixel export from raw – enlarged.

I don’t shoot at these speeds regularly. But I had used 800 once to catch a shot of Canarian musical instruments on the wall of a bar. CZ at 80mm, wide open at f/4.5, hand held 1/15th with SSS – not a recipe for a usable shot, but I have an interest in these (I own two lauds and a timple myself) and wanted a snap. I was amazed. The result at 800, in this low light, was similar to the Canon 450D at 800 and the Alpha 200 or 700 when used at 400. Not only that, no noise reduction was needed in ACR. It looked fine with zero luminance and zero chroma reduction. Slide chroma up to 50 then 100, and the image became as clean as I would expect from ISO 200.

At 1600 and 3200, however, the noise takes two quantum leaps in succession. The camera creates surprisingly good JPEGs with NR enabled, but both these and the raw files suffer from something which seems to plague Sony – the ‘grain’ is visually four times as big as the noise from the Pentax! It isn’t just noisier, the entire structure of the image becomes coarser and each ‘grain’ occupies a big cluster of output pixels, not a mere two or three. 3200 is for emergencies only, 1600 is less usable than the same setting on the 700 or 200.

What, then, is happening to make the lower ISO results exceptional? Presumably there is some clever processing on the sensor and in the raw file production. It does not appear to affect resolution – the detail capturing ability of the Alpha 350 goes beyond what would be expected from a mere 2 megapixels over the Alpha 700. While the Pentax’s rival 14.6 megapixel sensor will no doubt test out as even lower noise at ISO 100-800, in practice you can’t see any noise from either camera on most ISO 400 shots. Relative to the larger sensor size of the Alpha and Pentax, the Canon 450D’s 12.2 megapixel sensor has the same density as a 13.6 megapixel 1.5X factor DSLR. It’s got as good a 1600 result as the Pentax, but it’s worse then either the Pentax or Alpha at ISO 400.

Noise and detail must also be considered relative to the pixel count. Many of the shots I brought back for library filing would be a fair match to a Canon 1Ds MkII, helped by the Carl Zeiss 16-80mm. This is the important point to remember – the Alpha 350 is only as good as the lens you use.

For my money, Live View is most useful when composing shots that are difficult or impossible with a normal finder. Here, working at ground level.

And this is the detail the Alpha 350 can manage – once again, the raw file is exported to 5120 pixels wide using Adobe Camera Raw, the equivalent of shooting with a 17.5 megapixel camera.

My conclusion is that while the camera body I would prefer to use is the Alpha 700, I would dearly like to see an Alpha 750 fitted with the 14.2 megapixel CCD from the Alpha 350. The Live View I can take or leave; it makes some shots easier, but only because of the articulated viewing screen. Live View like the Canon 450D, 40D, Nikon D300, Olympus E-520 or Pentax K20D has no appeal to me outside the studio or macro bench; there is no advantage to holding up a DSLR and staring at a screen on the back. All my LV shots have been taken at waist-level, from a table, from the ground or overhead.

Which Alpha?

You’ll have realised that the answer is all four current choices, because they all do something different. The most distinctive choices are the 700 and 350. The 700 has the speed, accuracy and build needed for everyday professional work and the best overall balance of image qualities across the widest range of ISO settings. The 350 is best used in the range 100-400 ISO for subjects where detail rendering matters most. It offers Live View, but this is not the main reason to buy it.

The Alpha 200 and 300 have similar image quality (I am assuming that the 300 produces identical results in that respect). While the signal to noise ratio is higher than the original 100, there is only a marginal trade-off in colour and tone. It’s something which A100 owners may think is a small step away from ideal, but anyone coming from another system will be favourably impressed. This has happened before – Konica Minolta 7D and 5D owners felt the Alpha 100 moved a little away from the perfect colour and tone of those cameras!

Although you would expect the Alpha 700 and 200/300 to be more tolerant of exposure errors or contrasty conditions, the Alpha 350 turns out to be at least equal if not better, and there is no reason to shy clear of the high pixel count because you think it will have poor dynamic range or high noise at normal ISO settings. The larger files and slower shooting rate (2fps instead of 2.5fps) are the only negative things about choosing the 350 over the 200.

While the 350’s 14.2 megapixels don’t offer a big step from the A700’s 12.2, it is a significant increase on 10.2. If you really want to make larger prints, the A350 is well worth having. But Shirley, who uses our Alpha 200 right now, finds the smaller viewfinder, extra weight and different ergonomics of the A350 enough of a negative point to want to stick with the 200.

If you asked which one camera I would keep, it would be the Alpha 700 – but I’d find it more useful for my own work with that 14.2 megapixel sensor inside it!

– David Kilpatrick